Abstrakt
MR tractography of the lumbosacral plexus (LSP) is challenging due to the difficulty of acquiring high quality data and accurately estimating the neuronal tracts. We proposed an algorithm for an accurate visualization and assessment of the major LSP bundles using the segmentation of the cauda equina as seed points for the initial starting area for the fiber tracking algorithm.
Twenty-six healthy volunteers underwent MRI examinations on a 3T MR scanner using the phased array coils with optimized measurement protocols for diffusion-weighted images and coronal T2 weighted 3D short-term inversion recovery sampling perfection with application optimized contrast using varying flip angle evaluation sequences used for LSP fiber reconstruction and MR neurography (MRN). The fiber bundles reconstruction was optimized in terms of eliminating the muscle fibers contamination using the segmentation of cauda equina, the effects of the normalized quantitative anisotropy (NQA) and angular threshold on reconstruction of the LSP.
In this study, the NQA parameter has been used for fiber tracking instead of fractional anisotropy (FA) and the regions of interest positioning was precisely adjusted bilaterally and symmetrically in each individual subject. The diffusion data were processed in individual L3-S2 nerve fibers using the generalized Q-sampling imaging algorithm.
Data (mean FA, mean diffusivity, axial diffusivity and radial diffusivity, and normalized quantitative anisotropy) were statistically analyzed using the linear mixed-effects model. The MR neurography was performed in MedINRIA and post-processed using the maximum intensity projection method to demonstrate LSP tracts in multiple planes.
FA values significantly decreased towards the sacral region (P < .001); by contrast, mean diffusivity, axial diffusivity, radial diffusivity and NQA values significantly increased towards the sacral region (P < .001). Fiber tractography of the LSP was feasible in all examined subjects and closely corresponded with the nerves visible in the maximum intensity projection images of MR neurography.
Usage of NQA instead of FA in the proposed algorithm enabled better separation of muscle and nerve fibers. The presented algorithm yields a high quality reconstruction of the LSP bundles that may be helpful both in research and clinical practice.